Originally created by: andreas....@gmail.com
In rev 1748 it is not possible to change data in a grid if the grid data
was selected via a select statement in the "Query" register of a table.
After pressing ENTER on a selected field, changing the field and pressing
ENTER again the value of that field changes back to the original.
Tickets: #1123
Tickets: #1183
Tickets: #1393
Tickets: #1438
Tickets: #1634
Tickets: #443
View and moderate all "tickets Discussion" comments posted by this user
Mark all as spam, and block user from posting to "Tickets"
Originally posted by: a...@anse.de
That's not a bug, it's intentionally so, to avoid silent data loss bugs like
described in issue #324 .
Status: WontFix
Related
Tickets:
#324View and moderate all "tickets Discussion" comments posted by this user
Mark all as spam, and block user from posting to "Tickets"
Originally posted by: andreas....@gmail.com
Of cause it should not be possible to editors to allow editing data if a unique id
(unique index) was not part of the select satement. But that should be feasible if
there is one? And you can verife that and if here is one and the corresponding field
was part of the select statement you can allow editing?
View and moderate all "tickets Discussion" comments posted by this user
Mark all as spam, and block user from posting to "Tickets"
Originally posted by: a...@anse.de
A simple "SELECT bla FROM mytable" is only one of an endless number of different
combinations how a query can look like. Think for example of a JOIN:
SELECT tableA.col1 AS myalias FROM table1 AS tableA, table2 AS tableB WHERE
tableA.id = tB.id
How would you in such a case - technically spoken - compose a UPDATE statement if a
user updates the single field "myalias"? You'd say: No problem, split the SELECT
with some awesome SQL parser which detects everything - the real name of the
database, table, columns to updates in it, of course without having any security
issues in it so the user doesn't run the risk of breaking his data. We don't have
such a parser currently, which is the reason why the query result is blocked for
editing.
View and moderate all "tickets Discussion" comments posted by this user
Mark all as spam, and block user from posting to "Tickets"
Originally posted by: rosenfie...@gmail.com
MySQLd doesn't support editable results, which makes this difficult.
It does however return the name of the original table and original column with the
results of a SELECT. It also tells which columns are key columns.
HeidiSQL could do a SHOW KEYS on the original table after the user query has run,
find out if all key columns are also included in the results in the user query, and
allow UPDATEs via those if so.
This solves the key-finding part of the problem.
Another issue is with derived columns. They do not necessarily have to be just
aliases. They could just as well come from a concatenation, algebra, or whatever.
For example, if the query was "SELECT Id, Ships * 50 AS Containers FROM xxx", then
adding a new record and entering 500 in the 'Containers' column should actually
cause the update to post 10 into the original column.
I think this problem might be academic, since MySQLd *probably* does not return the
name of the original column, if the data was at all altered from the original column
data.
This just changes it into another problem though. What to do with columns for which
there are no "original table" and "original column" information? I suggest the
editor could just refuse to edit those columns, by refusing to go into edit mode.
All in all, I think it's doable, but would take some effort.
Labels: -Type-Defect Type-Enhancement
Status: Accepted
View and moderate all "tickets Discussion" comments posted by this user
Mark all as spam, and block user from posting to "Tickets"
Originally posted by: a...@anse.de
(No comment was entered for this change.)
Summary: Editable results in query tab
View and moderate all "tickets Discussion" comments posted by this user
Mark all as spam, and block user from posting to "Tickets"
Originally posted by: adamtho...@gmail.com
This functionality was a great feature and in many instances worked as required -
particularly on a simple select from where statement.
Could it not have a toggle on / off button on the control menu at the top of the
Heidi screen that is set as enabled=false where a join is used in the select?
View and moderate all "tickets Discussion" comments posted by this user
Mark all as spam, and block user from posting to "Tickets"
Originally posted by: a...@anse.de
Issue 1183 has been merged into this issue.
Related
Tickets:
#1183View and moderate all "tickets Discussion" comments posted by this user
Mark all as spam, and block user from posting to "Tickets"
Originally posted by: graveen....@gmail.com
Simply, i can understand the problems, but for simple query, it is really interested
to do a select, then clic 'n change.
I think a toggle button, not allowed by default, with a disclaimer could both warn
the possible trouble, and let the user a great flexibility.
View and moderate all "tickets Discussion" comments posted by this user
Mark all as spam, and block user from posting to "Tickets"
Originally posted by: rosenfie...@gmail.com
Issue 1212 has been merged into this issue.
Related
Tickets:
#1212View and moderate all "tickets Discussion" comments posted by this user
Mark all as spam, and block user from posting to "Tickets"
Originally posted by: a...@anse.de
Issue 1438 has been merged into this issue.
Related
Tickets:
#1438View and moderate all "tickets Discussion" comments posted by this user
Mark all as spam, and block user from posting to "Tickets"
Originally posted by: smujohn...@gmail.com
I agree with the people who think this feature should be implemented for simple
queries. I used it all the time in MySQL-Front (german one) when I was editing and
pasting simple field data for simple queries there were selected by an obvious key
index, etc. For complex queries and joins, I wouldn't expect it to work, much like
how MySQL-Front won't let you either.
View and moderate all "tickets Discussion" comments posted by this user
Mark all as spam, and block user from posting to "Tickets"
Originally posted by: a...@anse.de
Issue 1634 has been merged into this issue.
Related
Tickets:
#1634View and moderate all "tickets Discussion" comments posted by this user
Mark all as spam, and block user from posting to "Tickets"
Originally posted by: a...@anse.de
Fixed in [r3295]
Status: Fixed
Related
Commit: [r3295]